GCC 8 Might Pursue Better, More Modern Default Options

[phoronix]

Phoronix: GCC 8 Might Pursue Better, More Modern Default Options

Motivated by the 2017 GNU Tools Cauldron, an ARM developer is looking for feedback on improving the options enabled by default for the GCC 8 compiler...

http://www.phoronix.com/scan.php?page=news_item&px=GCC-8-Investigating-Defaults

[sdack]

Fly half around the world to meet in Prague,
drink Czechian beer for three days,
decide to omit the frame pointer on a few more platforms.

Worth it!

[tildearrow]

Typo:

Wilcon just posted this RFC message a short time ago

[duby229]

Yeah, fomit-frame-pointer actually exhibits some strange behavior that's probably not good. In some specific cases adding that flag can bypass a certain type of compile failure. But reading about what the flag does, I just don't understand how without the flag compile fails and with the flag compile finishes. It doesn't make logical sense in my head.

[sa666666]

Then that's a reason to research why the option seems to be doing something that doesn't make sense, not a justification for not using it at all.